Currently I am working for a Customer, in Media domain, which has webMethods based Integration layer forming its middleware tier. The Core Middleware Production webMethods environment comprises of about 15 different clusters. Each Cluster deals with specifc different Business functionality. (Example: There is a cluster to ‘Query Customer’s profile’ , a cluster to perform his ‘Credit Scoring’ and ‘Credit Card Evaluation’ etc) Each Cluster comprises of approx 30 webMethods Integration Server instances. Thus, including the fail-over servers, the count of total webMethods Integration Servers amount to approx 300 instances in Production. I would like to query the team and would be interested in hearing from anyone who would have witnessed and/or worked on such ginormous webMethods environment. I have certain specific queries. However, I would like to initiate a discussion with my basic curiosity. Awaiting your response.
Thanks in advance
wow,It looks pretty robust/critical environment of wM instances in place…What all other challenges you are facing?
300 instances is a huge set up by any means and I am now wondering what are the transaction volumes going through…have had dealt with 10s…would be interested to know more.
Mate i have worked with close to 480 IS’es distributed over a webMethods data farm spread across 3 physical locations using global and local layered hardwired xml based load balancers. guess this should be big enough
I have worked on a environment for two years with approx. 20 clusters of IS pairs (divided by different DCs) connecting two broker territories (which were in turn connected using gateways and were also distributed in the DCs).
One of the territories had some ~1000 single IS connected, with an excess of 20 connections each.
The biggest challenge was distributing the load on the brokers.
But apart from that, most inter-cluster communication was pretty straightforward.
What are the main challenges on your environment?